HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ichikawa, T. Articles by Marsh, W. Search for Related Content PubMed PubMed Citation Articles by Ichikawa, T. Articles by Marsh, W.

Fibrolamellar Hepatocellular Carcinoma: Imaging and Pathologic Findings in 31 Recent Cases¹

¹ From the Departments of Radiology (T.I., M.P.F., L.G.), Surgery (J.M., W.M.), and Pathology (M.N.), University of Pittsburgh Medical Center, 200 Lothrop St, Pittsburgh, PA 15213. Received December 30, 1998; revision requested February 8, 1999; revision received March 16; accepted June 9. Address reprint requests to M.P.F. (e-mail: federlemp@radserv.arad.upmc .edu).

	Abstract

TOP Abstract Introduction MATERIALS AND METHODS RESULTS DISCUSSION References

 
PURPOSE: To review characteristic findings of fibrolamellar hepatocellular carcinoma (HCC) at computed tomography (CT) and magnetic resonance (MR) imaging.

MATERIALS AND METHODS: The authors retrospectively reviewed the clinical, pathologic, and preoperative imaging findings in 31 patients with histologically proved fibrolamellar HCC. Dynamic contrast material–enhanced CT of the liver was performed in 31 patients, helical multiphase CT in 21, and MR imaging in 11. Complete resection was performed in 17 patients, and imaging-pathologic correlation was performed.

RESULTS: Large tumors (mean diameter, 13 cm) were depicted at CT and MR in all cases. At CT, the margins of the tumors were well defined in 24 (77%) of 31 cases, calcifications were depicted in 21 (68%), a central scar in 22 (71%), and abdominal lymphadenopathy in 20 (65%). In 20 (80%) of 25 cases with hepatic arterial phase CT images, all tumors were heterogeneous and depicted areas of hypervascularity. At MR imaging, tumors were hypointense to liver on TI-weighted images (n = 11) and hyperintense to liver on T2-weighted images (n = 10). Calcification was not depicted on MR images, but a central scar was depicted as hypointense to surrounding tumor in nine cases.

CONCLUSION: CT and MR images demonstrate characteristic features that may allow confident diagnosis of fibrolamellar HCC.

Index terms: Liver neoplasms, CT, 761.12111, 761.12112, 761.12113, 761.12114, 761.12115 • Liver neoplasms, diagnosis, 761.324 • Liver neoplasms, MR, 761.121411, 761.121412, 761.12143 • Magnetic resonance (MR), contrast agents, 761.12143

	Introduction

TOP Abstract Introduction MATERIALS AND METHODS RESULTS DISCUSSION References

 
Fibrolamellar hepatocellular carcinoma (HCC) is an uncommon tumor that has distinctive clinical, histologic, and prognostic features. Among these are its frequency among young patients with no history of cirrhosis or chronic liver disease, absence of serum tumor markers, increased chance of resectability for cure, and prolonged survival compared with that in patients with conventional HCC (1–6). Although it is important to distinguish fibrolamellar HCC from conventional HCC, it is equally important to distinguish it from certain benign liver masses, especially liver cell adenoma and focal nodular hyperplasia (FNH).

In prior investigations, the difficulty of distinguishing fibrolamellar HCC from other hypervascular liver masses on the basis of cross-sectional imaging studies alone has been emphasized (7–17). Percutaneous biopsy and histopathologic examination may also not always allow definitive diagnosis. As a result, patients may be denied potentially curative resection or may undergo unnecessary resection on the basis of a belief that fibrolamellar HCC cannot be distinguished reliably from conventional HCC, adenoma, or FNH.

Almost all reports of the radiologic features of fibrolamellar HCC consist of case reports and small series (7–17). To our knowledge, there has been no published study of a large series of cases of fibrolamellar HCC that have been investigated with helical multiphase contrast material–enhanced computed tomography (CT) or newer magnetic resonance (MR) techniques. The purpose of this study was to perform such an analysis, in the hope of deriving imaging characteristics that might allow more confident diagnosis and effective management of fibrolamellar HCC.

	MATERIALS AND METHODS

TOP Abstract Introduction MATERIALS AND METHODS RESULTS DISCUSSION References

 
We reviewed the pathology registry at our institution from 1989 to 1998 and identified 76 patients with a histologic diagnosis of fibrolamellar HCC. Three patients were excluded because only small foci of fibrolamellar HCC were identified. Among the 73 remaining patients, 40 had undergone abdominal CT, with or without other imaging studies, followed by surgery or other therapy at this institution. Nine patients were eliminated because the pertinent CT scans could not be located. Finally, we retrospectively reviewed the clinical, pathologic, and preoperative imaging features in the remaining 31 patients, who had complete records.

There were 16 men and 15 women, aged 15–56 years (mean, 28 years). Complete tumor resection was performed in 17 of the 31 cases by means of partial hepatectomy (n = 13) or orthotopic liver transplantation (n = 4). In the remaining 14 cases, pathologic specimens were obtained at exploratory laparotomy (n = 6) or percutaneous needle biopsy (n = 8). No patients had underlying hepatic cirrhosis, and none had elevated serum -fetoprotein levels.

Histopathologic confirmation of fibrolamellar HCC was based on demonstration of large polygonal eosinophilic malignant hepatocytes dispersed throughout a fibrolamellar stroma in patients without underlying cirrhosis.

All 31 patients had undergone abdominal CT evaluation, and 11 had also undergone hepatic MR examinations. All CT examinations included nonenhanced and contrast-enhanced imaging through the liver. Ten patients underwent conventional transverse (nonhelical) CT, and 21 underwent helical multiphase examinations including both hepatic arterial phase and portal venous phase imaging through the liver, with delays of 25 seconds and 60–70 seconds, respectively, after initiation of infusion of the bolus of intravenous contrast material. A scanning delay of 60 seconds was used for conventional transverse CT. On average, patients received 150 mL of 60% iodinated contrast medium (iothalamate meglumine, Conray 60, Mallinckrodt Medical, St Louis, Mo, or ioversol, Optiray 320, Mallinckrodt Medical). At conventional CT, intravenous contrast material was administered at a rate of 2–3 mL/sec. At helical multiphase CT, the contrast agent was administered at a rate of 4 or 5 mL/sec with a power injector (model OP 100; Medrad, Pittsburgh, Pa). In addition, nine patients also underwent delayed-phase imaging through the liver, 10–20 minutes after initiation of the contrast medium administration. Section thickness was 5–7 mm for conventional and helical imaging. All CT studies were performed with CT HiLight Advantage or CT HiSpeed Advantage scanners (GE Medical Systems, Milwaukee, Wis).

Eleven of the 31 patients underwent MR imaging with a T1-weighted spin-echo sequence (repetition time msec/echo time msec = 140–700/12–20). In addition, T2-weighted spin-echo imaging (4,000–9,230/70–140; echo train length, eight or 16) was performed in four and fast spin-echo imaging (4,000–6,000/102) in seven. Fat-suppression techniques were not used. T1-weighted imaging was repeated in nine patients after contrast material administration. Five patients received gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) at a dose of 0.1 mmol per kg of body weight, and four underwent biphasic MR imaging during the hepatic arterial (25-second delay) and portal venous (60–70-second delay) phases. Three of these four patients underwent MR imaging after intravenous administration of mangafodipir trisodium (Teslascan; Nycomed-Amersham, Princeton, NJ) at a dose of 5 µmol/kg. One patient underwent MR imaging before and after intravenous administration of gadobenate dimeglumine (Gd-BOPTA; Bracco, Milan, Italy) at a dose of 0.1 mmol/kg. The T1- and T2-weighted sequences were similar regardless of which contrast agent was used. All MR studies were performed with Signa 1.5-T imagers (GE Medical Systems) with a variety of software upgrades that continuously evolved over the period of the study.

All imaging studies were interpreted retrospectively and independently by two radiologists (T.I., L.G.) without knowledge of specific clinical and pathologic findings but with knowledge of the diagnosis of fibrolamellar HCC. CT and MR studies in the same patient were reviewed together. Specific and detailed imaging criteria were sought and recorded for each patient and each study, with use of imaging studies completed only before initiation of any therapy. These findings were used to correlate the imaging and gross pathologic features of these tumors. In cases of interobserver disagreement, final decisions were reached by consensus.

Imaging Criteria
CT findings were tabulated and included the bidirectional maximum diameter of the tumor and its margins (well defined or ill defined) and surface (smooth or lobulated). A tumor capsule was defined as a curvilinear border that surrounded the tumor and had a distinct attenuation difference. The overall attenuation of the tumor was defined relative to the liver during the same phase (hepatic arterial or portal venous) of imaging. The presence, size, and distribution of calcifications were recorded. Nonenhancing areas with attenuation similar to that of gallbladder contents were regarded as cystlike, representing necrosis or nonacute hemorrhage. Hemorrhage was regarded as amorphous fluid with attenuation higher than that of the nonenhanced liver. Visual, subjective criteria were used rather than region-of-interest measurements.

Upper abdominal lymphadenopathy was diagnosed when ovoid or rounded extravisceral masses were identified that were 2 cm or more in diameter and had attenuation less than or equal to that of skeletal muscle.

MR findings also included measures of tumor size, margination, and encapsulation. Signal intensity of the tumor was defined relative to liver and other structures on the same image. On T2-weighted spin-echo images, marked hyperintensity was defined as signal intensity more than that of liver, similar to that of fat, but less than that of cerebrospinal fluid. Fluid or cystlike components were defined as having signal intensity similar to that of cerebrospinal fluid on both T1- and T2-weighted images.

Gross Pathologic Criteria
The 17 cases in which complete resection of the fibrolamellar HCC was performed and the 14 cases with biopsy specimens were reviewed retrospectively by one pathologist (M.N.) with expertise in hepatic pathology. Microscopic sections were reviewed for confirmation of the diagnosis of fibrolamellar HCC. The gross specimens were evaluated for features similar to those defined by imaging characteristics, namely tumor size, margins, and encapsulation, and the presence and extent of a central scar, fibrous bands or septa, necrosis, calcification, or hemorrhage.

	RESULTS

TOP Abstract Introduction MATERIALS AND METHODS RESULTS DISCUSSION References

 
CT Studies
CT demonstrated an obvious tumor in all 31 cases, with a mean maximum diameter of 13 cm (range, 3–27 cm) (Figs 1–5). The margins of the tumor were well defined in 24 cases (77%) (Figs 2–5) and ill defined in seven cases. Among the well-defined tumors, the surface was lobulated in 20 cases (83%) (Figs 1–5) and smooth in four. The tumors were located predominantly in the left lobe (20 of 31 cases [65%]) with the lateral segment and medial segment involved predominantly in 14 and six cases, respectively (Figs 1, 3–5). The right lobe was the apparent epicenter of the tumor in 11 cases (35%). However, more than half the cases (17 of 31 [55%]) had involvement of three or more hepatic segments. Only four patients (13%) had involvement of a single segment, whereas 10 (32%) had involvement of two segments.

View larger version (128K): [in this window] [in a new window]   Figure 1a. Large exophytic fibrolamellar HCC arising from the lateral segment in a 27-year-old man. (a) Series of transverse nonenhanced CT sections demonstrate the large mass (straight solid arrows) hypoattenuating to the liver, with calcification (curved arrow) present in the central scar (open arrow). The stomach (S) is compressed by the tumor. (b) Hepatic arterial phase transverse CT images demonstrate heterogeneous enhancement of the tumor (arrows) with parts of the mass slightly hyperattenuating to the normal right lobe. The medial (ms) and lateral (ls) segments of the left lobe that are free of tumor are enhanced to an abnormal degree, implying increased arterial flow and decreased portal venous flow (transient hepatic attenuation difference). (c) Portal venous transverse CT images demonstrate the heterogeneous tumor (straight arrows) and compression or thrombosis of the left portal vein (curved arrow), which accounts for the transient hepatic attenuation difference. (d) Other portal venous transverse CT images demonstrate extensive upper abdominal lymphadenopathy (ln). (e) Photograph of the gross pathologic specimen of the surface of the resected lateral segment of the liver demonstrates the lobulated contour (arrows) of the tumor. (f) Photograph of the gross pathologic specimen of the cross section through the tumor reveals an irregular central scar (open arrows) and stellate fibrous septa. The surrounding tumor is fleshy and has a variegated appearance owing to bile staining. A focus of necrosis (black arrow) is seen in the tumor periphery. The tumor borders are grossly circumscribed but lobulated and irregular (curved arrows).

View larger version (143K): [in this window] [in a new window]   Figure 1b. Large exophytic fibrolamellar HCC arising from the lateral segment in a 27-year-old man. (a) Series of transverse nonenhanced CT sections demonstrate the large mass (straight solid arrows) hypoattenuating to the liver, with calcification (curved arrow) present in the central scar (open arrow). The stomach (S) is compressed by the tumor. (b) Hepatic arterial phase transverse CT images demonstrate heterogeneous enhancement of the tumor (arrows) with parts of the mass slightly hyperattenuating to the normal right lobe. The medial (ms) and lateral (ls) segments of the left lobe that are free of tumor are enhanced to an abnormal degree, implying increased arterial flow and decreased portal venous flow (transient hepatic attenuation difference). (c) Portal venous transverse CT images demonstrate the heterogeneous tumor (straight arrows) and compression or thrombosis of the left portal vein (curved arrow), which accounts for the transient hepatic attenuation difference. (d) Other portal venous transverse CT images demonstrate extensive upper abdominal lymphadenopathy (ln). (e) Photograph of the gross pathologic specimen of the surface of the resected lateral segment of the liver demonstrates the lobulated contour (arrows) of the tumor. (f) Photograph of the gross pathologic specimen of the cross section through the tumor reveals an irregular central scar (open arrows) and stellate fibrous septa. The surrounding tumor is fleshy and has a variegated appearance owing to bile staining. A focus of necrosis (black arrow) is seen in the tumor periphery. The tumor borders are grossly circumscribed but lobulated and irregular (curved arrows).

View larger version (135K): [in this window] [in a new window]   Figure 1c. Large exophytic fibrolamellar HCC arising from the lateral segment in a 27-year-old man. (a) Series of transverse nonenhanced CT sections demonstrate the large mass (straight solid arrows) hypoattenuating to the liver, with calcification (curved arrow) present in the central scar (open arrow). The stomach (S) is compressed by the tumor. (b) Hepatic arterial phase transverse CT images demonstrate heterogeneous enhancement of the tumor (arrows) with parts of the mass slightly hyperattenuating to the normal right lobe. The medial (ms) and lateral (ls) segments of the left lobe that are free of tumor are enhanced to an abnormal degree, implying increased arterial flow and decreased portal venous flow (transient hepatic attenuation difference). (c) Portal venous transverse CT images demonstrate the heterogeneous tumor (straight arrows) and compression or thrombosis of the left portal vein (curved arrow), which accounts for the transient hepatic attenuation difference. (d) Other portal venous transverse CT images demonstrate extensive upper abdominal lymphadenopathy (ln). (e) Photograph of the gross pathologic specimen of the surface of the resected lateral segment of the liver demonstrates the lobulated contour (arrows) of the tumor. (f) Photograph of the gross pathologic specimen of the cross section through the tumor reveals an irregular central scar (open arrows) and stellate fibrous septa. The surrounding tumor is fleshy and has a variegated appearance owing to bile staining. A focus of necrosis (black arrow) is seen in the tumor periphery. The tumor borders are grossly circumscribed but lobulated and irregular (curved arrows).

View larger version (143K): [in this window] [in a new window]   Figure 1d. Large exophytic fibrolamellar HCC arising from the lateral segment in a 27-year-old man. (a) Series of transverse nonenhanced CT sections demonstrate the large mass (straight solid arrows) hypoattenuating to the liver, with calcification (curved arrow) present in the central scar (open arrow). The stomach (S) is compressed by the tumor. (b) Hepatic arterial phase transverse CT images demonstrate heterogeneous enhancement of the tumor (arrows) with parts of the mass slightly hyperattenuating to the normal right lobe. The medial (ms) and lateral (ls) segments of the left lobe that are free of tumor are enhanced to an abnormal degree, implying increased arterial flow and decreased portal venous flow (transient hepatic attenuation difference). (c) Portal venous transverse CT images demonstrate the heterogeneous tumor (straight arrows) and compression or thrombosis of the left portal vein (curved arrow), which accounts for the transient hepatic attenuation difference. (d) Other portal venous transverse CT images demonstrate extensive upper abdominal lymphadenopathy (ln). (e) Photograph of the gross pathologic specimen of the surface of the resected lateral segment of the liver demonstrates the lobulated contour (arrows) of the tumor. (f) Photograph of the gross pathologic specimen of the cross section through the tumor reveals an irregular central scar (open arrows) and stellate fibrous septa. The surrounding tumor is fleshy and has a variegated appearance owing to bile staining. A focus of necrosis (black arrow) is seen in the tumor periphery. The tumor borders are grossly circumscribed but lobulated and irregular (curved arrows).

View larger version (158K): [in this window] [in a new window]   Figure 1e. Large exophytic fibrolamellar HCC arising from the lateral segment in a 27-year-old man. (a) Series of transverse nonenhanced CT sections demonstrate the large mass (straight solid arrows) hypoattenuating to the liver, with calcification (curved arrow) present in the central scar (open arrow). The stomach (S) is compressed by the tumor. (b) Hepatic arterial phase transverse CT images demonstrate heterogeneous enhancement of the tumor (arrows) with parts of the mass slightly hyperattenuating to the normal right lobe. The medial (ms) and lateral (ls) segments of the left lobe that are free of tumor are enhanced to an abnormal degree, implying increased arterial flow and decreased portal venous flow (transient hepatic attenuation difference). (c) Portal venous transverse CT images demonstrate the heterogeneous tumor (straight arrows) and compression or thrombosis of the left portal vein (curved arrow), which accounts for the transient hepatic attenuation difference. (d) Other portal venous transverse CT images demonstrate extensive upper abdominal lymphadenopathy (ln). (e) Photograph of the gross pathologic specimen of the surface of the resected lateral segment of the liver demonstrates the lobulated contour (arrows) of the tumor. (f) Photograph of the gross pathologic specimen of the cross section through the tumor reveals an irregular central scar (open arrows) and stellate fibrous septa. The surrounding tumor is fleshy and has a variegated appearance owing to bile staining. A focus of necrosis (black arrow) is seen in the tumor periphery. The tumor borders are grossly circumscribed but lobulated and irregular (curved arrows).

View larger version (145K): [in this window] [in a new window]   Figure 1f. Large exophytic fibrolamellar HCC arising from the lateral segment in a 27-year-old man. (a) Series of transverse nonenhanced CT sections demonstrate the large mass (straight solid arrows) hypoattenuating to the liver, with calcification (curved arrow) present in the central scar (open arrow). The stomach (S) is compressed by the tumor. (b) Hepatic arterial phase transverse CT images demonstrate heterogeneous enhancement of the tumor (arrows) with parts of the mass slightly hyperattenuating to the normal right lobe. The medial (ms) and lateral (ls) segments of the left lobe that are free of tumor are enhanced to an abnormal degree, implying increased arterial flow and decreased portal venous flow (transient hepatic attenuation difference). (c) Portal venous transverse CT images demonstrate the heterogeneous tumor (straight arrows) and compression or thrombosis of the left portal vein (curved arrow), which accounts for the transient hepatic attenuation difference. (d) Other portal venous transverse CT images demonstrate extensive upper abdominal lymphadenopathy (ln). (e) Photograph of the gross pathologic specimen of the surface of the resected lateral segment of the liver demonstrates the lobulated contour (arrows) of the tumor. (f) Photograph of the gross pathologic specimen of the cross section through the tumor reveals an irregular central scar (open arrows) and stellate fibrous septa. The surrounding tumor is fleshy and has a variegated appearance owing to bile staining. A focus of necrosis (black arrow) is seen in the tumor periphery. The tumor borders are grossly circumscribed but lobulated and irregular (curved arrows).

View larger version (118K): [in this window] [in a new window]   Figure 2a. Large fibrolamellar HCC mainly in the right lobe in a 25-year-old woman. (a) Transverse T1-weighted gradient-echo MR image (120/4.2 with 90° flip angle) shows the large mass (arrows), which is predominantly hypointense to normal liver (Liv). (b) Hepatic arterial transverse T1-weighted gradient-echo image (120/4.2) obtained during bolus infusion of gadopentetate dimeglumine demonstrates marked heterogeneity of the tumor, hyperintensity of some portions, eccentric scar and fibrous septa (open arrows), and a nonenhancing capsule (solid arrows) that surrounds most of the tumor. (c) Transverse T2-weighted MR image (9,230/99; echo train length, 16) demonstrates the hypointense areas of fibrous tissue (open arrows) and scattered hyperintense foci (curved arrow) that represent areas of necrosis. (d) Photograph of the pathologic specimen of the cut surface demonstrates an asymmetric multinodular tumor that consists of several masses linked by fibrous scar (open arrows). The largest mass shows multiple cystlike areas of necrosis (curved arrow). A fibrous capsule is also evident around some of the tumor (straight solid arrows).

View larger version (131K): [in this window] [in a new window]   Figure 2b. Large fibrolamellar HCC mainly in the right lobe in a 25-year-old woman. (a) Transverse T1-weighted gradient-echo MR image (120/4.2 with 90° flip angle) shows the large mass (arrows), which is predominantly hypointense to normal liver (Liv). (b) Hepatic arterial transverse T1-weighted gradient-echo image (120/4.2) obtained during bolus infusion of gadopentetate dimeglumine demonstrates marked heterogeneity of the tumor, hyperintensity of some portions, eccentric scar and fibrous septa (open arrows), and a nonenhancing capsule (solid arrows) that surrounds most of the tumor. (c) Transverse T2-weighted MR image (9,230/99; echo train length, 16) demonstrates the hypointense areas of fibrous tissue (open arrows) and scattered hyperintense foci (curved arrow) that represent areas of necrosis. (d) Photograph of the pathologic specimen of the cut surface demonstrates an asymmetric multinodular tumor that consists of several masses linked by fibrous scar (open arrows). The largest mass shows multiple cystlike areas of necrosis (curved arrow). A fibrous capsule is also evident around some of the tumor (straight solid arrows).

View larger version (133K): [in this window] [in a new window]   Figure 2c. Large fibrolamellar HCC mainly in the right lobe in a 25-year-old woman. (a) Transverse T1-weighted gradient-echo MR image (120/4.2 with 90° flip angle) shows the large mass (arrows), which is predominantly hypointense to normal liver (Liv). (b) Hepatic arterial transverse T1-weighted gradient-echo image (120/4.2) obtained during bolus infusion of gadopentetate dimeglumine demonstrates marked heterogeneity of the tumor, hyperintensity of some portions, eccentric scar and fibrous septa (open arrows), and a nonenhancing capsule (solid arrows) that surrounds most of the tumor. (c) Transverse T2-weighted MR image (9,230/99; echo train length, 16) demonstrates the hypointense areas of fibrous tissue (open arrows) and scattered hyperintense foci (curved arrow) that represent areas of necrosis. (d) Photograph of the pathologic specimen of the cut surface demonstrates an asymmetric multinodular tumor that consists of several masses linked by fibrous scar (open arrows). The largest mass shows multiple cystlike areas of necrosis (curved arrow). A fibrous capsule is also evident around some of the tumor (straight solid arrows).

View larger version (112K): [in this window] [in a new window]   Figure 2d. Large fibrolamellar HCC mainly in the right lobe in a 25-year-old woman. (a) Transverse T1-weighted gradient-echo MR image (120/4.2 with 90° flip angle) shows the large mass (arrows), which is predominantly hypointense to normal liver (Liv). (b) Hepatic arterial transverse T1-weighted gradient-echo image (120/4.2) obtained during bolus infusion of gadopentetate dimeglumine demonstrates marked heterogeneity of the tumor, hyperintensity of some portions, eccentric scar and fibrous septa (open arrows), and a nonenhancing capsule (solid arrows) that surrounds most of the tumor. (c) Transverse T2-weighted MR image (9,230/99; echo train length, 16) demonstrates the hypointense areas of fibrous tissue (open arrows) and scattered hyperintense foci (curved arrow) that represent areas of necrosis. (d) Photograph of the pathologic specimen of the cut surface demonstrates an asymmetric multinodular tumor that consists of several masses linked by fibrous scar (open arrows). The largest mass shows multiple cystlike areas of necrosis (curved arrow). A fibrous capsule is also evident around some of the tumor (straight solid arrows).

View larger version (142K): [in this window] [in a new window]   Figure 3a. A large central fibrolamellar HCC in a 37-year-old man. (a) Nonenhanced transverse CT scan clearly demonstrates the hypoattenuating tumor (straight arrows) and shows one of several calcifications (curved arrow). (b) Portal venous contrast-enhanced transverse CT scan demonstrates heterogeneity and may suggest a central scar (arrow), but most of the tumor is isoattenuating to liver and poorly defined. (c) Transverse T2-weighted MR image (4,433/85) better demonstrates the tumor margins (solid arrows) and some morphologic characteristics such as central and radiating fibrous tissue (open arrow), but the calcifications are not evident.

View larger version (133K): [in this window] [in a new window]   Figure 3b. A large central fibrolamellar HCC in a 37-year-old man. (a) Nonenhanced transverse CT scan clearly demonstrates the hypoattenuating tumor (straight arrows) and shows one of several calcifications (curved arrow). (b) Portal venous contrast-enhanced transverse CT scan demonstrates heterogeneity and may suggest a central scar (arrow), but most of the tumor is isoattenuating to liver and poorly defined. (c) Transverse T2-weighted MR image (4,433/85) better demonstrates the tumor margins (solid arrows) and some morphologic characteristics such as central and radiating fibrous tissue (open arrow), but the calcifications are not evident.

View larger version (133K): [in this window] [in a new window]   Figure 3c. A large central fibrolamellar HCC in a 37-year-old man. (a) Nonenhanced transverse CT scan clearly demonstrates the hypoattenuating tumor (straight arrows) and shows one of several calcifications (curved arrow). (b) Portal venous contrast-enhanced transverse CT scan demonstrates heterogeneity and may suggest a central scar (arrow), but most of the tumor is isoattenuating to liver and poorly defined. (c) Transverse T2-weighted MR image (4,433/85) better demonstrates the tumor margins (solid arrows) and some morphologic characteristics such as central and radiating fibrous tissue (open arrow), but the calcifications are not evident.

View larger version (130K): [in this window] [in a new window]   Figure 4a. Left-lobe fibrolamellar HCC in a 19-year-old man. (a) Non-enhanced transverse CT scan shows calcification (curved arrow) within the hypoattenuating tumor (straight arrows). (b) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). (c) Ten-minute delayed transverse CT scan demonstrates subtle areas of hyperattenuation that represent fibrous tissue within the central scar, radiating septa, and capsule (open arrows). Curved arrow = calcification. (d) Transverse T2-weighted MR image (5,000/105) also demonstrates the central scar and septa (open arrow). The tumor itself (straight arrows) is nearly isointense to liver (the only such case in our series). (e) Photograph of the cut surface of the gross pathologic specimen shows a large tumor with eccentric and central scars (open arrows) and radiating septa. The mass has an irregular lobulated pushing margin (solid arrows) and a variegated appearance with areas of bile staining.

View larger version (139K): [in this window] [in a new window]   Figure 4b. Left-lobe fibrolamellar HCC in a 19-year-old man. (a) Non-enhanced transverse CT scan shows calcification (curved arrow) within the hypoattenuating tumor (straight arrows). (b) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). (c) Ten-minute delayed transverse CT scan demonstrates subtle areas of hyperattenuation that represent fibrous tissue within the central scar, radiating septa, and capsule (open arrows). Curved arrow = calcification. (d) Transverse T2-weighted MR image (5,000/105) also demonstrates the central scar and septa (open arrow). The tumor itself (straight arrows) is nearly isointense to liver (the only such case in our series). (e) Photograph of the cut surface of the gross pathologic specimen shows a large tumor with eccentric and central scars (open arrows) and radiating septa. The mass has an irregular lobulated pushing margin (solid arrows) and a variegated appearance with areas of bile staining.

View larger version (121K): [in this window] [in a new window]   Figure 4c. Left-lobe fibrolamellar HCC in a 19-year-old man. (a) Non-enhanced transverse CT scan shows calcification (curved arrow) within the hypoattenuating tumor (straight arrows). (b) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). (c) Ten-minute delayed transverse CT scan demonstrates subtle areas of hyperattenuation that represent fibrous tissue within the central scar, radiating septa, and capsule (open arrows). Curved arrow = calcification. (d) Transverse T2-weighted MR image (5,000/105) also demonstrates the central scar and septa (open arrow). The tumor itself (straight arrows) is nearly isointense to liver (the only such case in our series). (e) Photograph of the cut surface of the gross pathologic specimen shows a large tumor with eccentric and central scars (open arrows) and radiating septa. The mass has an irregular lobulated pushing margin (solid arrows) and a variegated appearance with areas of bile staining.

View larger version (135K): [in this window] [in a new window]   Figure 4d. Left-lobe fibrolamellar HCC in a 19-year-old man. (a) Non-enhanced transverse CT scan shows calcification (curved arrow) within the hypoattenuating tumor (straight arrows). (b) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). (c) Ten-minute delayed transverse CT scan demonstrates subtle areas of hyperattenuation that represent fibrous tissue within the central scar, radiating septa, and capsule (open arrows). Curved arrow = calcification. (d) Transverse T2-weighted MR image (5,000/105) also demonstrates the central scar and septa (open arrow). The tumor itself (straight arrows) is nearly isointense to liver (the only such case in our series). (e) Photograph of the cut surface of the gross pathologic specimen shows a large tumor with eccentric and central scars (open arrows) and radiating septa. The mass has an irregular lobulated pushing margin (solid arrows) and a variegated appearance with areas of bile staining.

View larger version (128K): [in this window] [in a new window]   Figure 4e. Left-lobe fibrolamellar HCC in a 19-year-old man. (a) Non-enhanced transverse CT scan shows calcification (curved arrow) within the hypoattenuating tumor (straight arrows). (b) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). (c) Ten-minute delayed transverse CT scan demonstrates subtle areas of hyperattenuation that represent fibrous tissue within the central scar, radiating septa, and capsule (open arrows). Curved arrow = calcification. (d) Transverse T2-weighted MR image (5,000/105) also demonstrates the central scar and septa (open arrow). The tumor itself (straight arrows) is nearly isointense to liver (the only such case in our series). (e) Photograph of the cut surface of the gross pathologic specimen shows a large tumor with eccentric and central scars (open arrows) and radiating septa. The mass has an irregular lobulated pushing margin (solid arrows) and a variegated appearance with areas of bile staining.

View larger version (136K): [in this window] [in a new window]   Figure 5a. Images in a 22-year-old woman with left-lobe fibrolamellar HCC. (a) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). On the portal venous contrast-enhanced CT images (not shown), the tumor was heterogeneous but nearly isoattenuating to the liver. (b) Ten-minute delayed transverse CT image demonstrates central and radiating areas of hyperattenuation (arrow) that indicate fibrous stroma. (c) Photograph of the gross pathologic specimen of the cut surface clearly demonstrates the large central fibrous scar and radiating septa (open arrow). Solid arrows indicate tumor margins.

View larger version (120K): [in this window] [in a new window]   Figure 5b. Images in a 22-year-old woman with left-lobe fibrolamellar HCC. (a) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). On the portal venous contrast-enhanced CT images (not shown), the tumor was heterogeneous but nearly isoattenuating to the liver. (b) Ten-minute delayed transverse CT image demonstrates central and radiating areas of hyperattenuation (arrow) that indicate fibrous stroma. (c) Photograph of the gross pathologic specimen of the cut surface clearly demonstrates the large central fibrous scar and radiating septa (open arrow). Solid arrows indicate tumor margins.

View larger version (131K): [in this window] [in a new window]   Figure 5c. Images in a 22-year-old woman with left-lobe fibrolamellar HCC. (a) Hepatic arterial contrast-enhanced transverse CT scan shows heterogeneous hypervascularity within the tumor (arrows). On the portal venous contrast-enhanced CT images (not shown), the tumor was heterogeneous but nearly isoattenuating to the liver. (b) Ten-minute delayed transverse CT image demonstrates central and radiating areas of hyperattenuation (arrow) that indicate fibrous stroma. (c) Photograph of the gross pathologic specimen of the cut surface clearly demonstrates the large central fibrous scar and radiating septa (open arrow). Solid arrows indicate tumor margins.

Cystlike or necrotic areas were detected in 20 of 31 cases (65%), which occupied less than half the tumor volume in 17 cases and more than half in only three (Figs 1, 2). Hyperattenuating areas compatible with hemorrhage were noted in only one case within the tumor itself, in which a subcapsular hepatic hematoma was confirmed at surgery. No tumor had CT evidence of a fat component.

On nonenhanced CT images, the fibrolamellar HCC was predominantly hypoattenuating compared with the liver in 29 of 31 cases (94%) (Figs 1, 3, 4) and isoattenuating in two cases. Almost all tumors were heterogeneous on nonenhanced CT images (28 of 31 [90%]). The three tumors characterized as homogeneous were among the smaller masses (diameter, 3, 5, and 8 cm).

Twenty-one patients underwent helical dual-phase imaging of the liver during rapid bolus infusion of contrast material. In addition, four of the 10 cases with conventional CT scans had liver images predominantly in the arterial dominant phase. On hepatic arterial phase images, the fibrolamellar HCC was predominantly but heterogeneously hyperattenuating to liver in 20 of 25 cases (80%) (Figs 1, 4, 5). Even in the four cases characterized as hypoattenuating, obvious brightly enhancing tumor vessels were identified within the tumor on the hepatic arterial phase images.

Portal venous phase images of the liver were obtained in 21 cases with helical dual-phase imaging and in an additional four cases with conventional transverse CT. Relative to enhanced liver, the fibrolamellar HCC appeared isoattenuating in 12 of 25 cases (48%), hyperattenuating in four (16%), and hypoattenuating in nine (36%).

On delayed-phase images in nine patients, parts of the nonnecrotic portions of the tumor increased in attenuation relative to liver. The fibrolamellar HCC was judged predominantly hyperattenuating in two of the nine cases (22%), isoattenuating in two (22%), and hypoattenuating in five (56%).

CT demonstrated a central scar in 22 of 31 cases (71%) (Figs 4, 5) that was predominantly stellate in 18 (82%) and amorphous in four. The size of the central scar was described as large in six tumors, moderate in 15, and small in one. The central scar had radial septal bands of density in 21 of the 22 cases (95%) that extended toward the periphery of the tumor (Figs 4, 5). In a comparison of the central scar with the surrounding tumor tissue, the scar was hypoattenuating to tumor on both nonenhanced and contrast-enhanced scans. Even at helical imaging on arterial and portal venous phase images, the scar showed minimal enhancement in all but one case. On delayed-phase images, the central scar became hyperattenuating to tumor in five of nine cases (56%) (Figs 4, 5) and remained hypoattenuating in four. Calcification within the central scar was identified in 20 of 22 cases (91%) with scars (Figs 1, 3, 4). The frequent and characteristic CT findings, defined as occurrence in more than 50% of cases, are summarized in Table 1.

MR Imaging Studies
MR imaging demonstrated a large tumor in all 11 cases, and each was hypointense to liver on T1-weighted images (Fig 2). On T2-weighted images, the tumor was hyperintense to liver in 10 of 11 cases (91%), moderate in nine, and markedly hyperintense in one (Figs 2, 3). The tumor was isointense to liver in one case. Calcification was seldom suggested by hypointense spots on both T1- and T2-weighted images. In four cases, an incomplete tumor capsule was demonstrated that was hypointense on both T1- and T2-weighted images (Fig 2). Cystlike or necrotic areas that composed less than 25% of the tumor were noted in four cases (36%); they appeared hypointense on T1-weighted images and markedly hyperintense on T2-weighted images. No tumor had MR findings suggestive of a fat component.

On mangafodipir-enhanced MR images (n = 3), the tumor showed no enhancement in two cases and definite enhancement isointense to enhanced liver in one case. In the only patient with a gadobenate dimeglumine–enhanced MR image, the tumor did not enhance and became relatively more hypointense to liver on T1-weighted images.

A central scar was demonstrated in nine of 11 tumors (82%) as a stellate region hypointense to surrounding tumor and liver on both T1- and T2-weighted images (Figs 2–4). CT demonstrated a scar in only six of these nine cases. T2-weighted images also demonstrated hypointense septa radiating toward the tumor periphery and depicted this feature more clearly than did CT in seven of nine cases (78%) (Figs 3, 4). As at CT, the central scar showed minimal or no enhancement on hepatic arterial or portal venous MR images, which made it relatively more hypointense to surrounding tumor. The three cases with delayed-phase gadopentetate dimeglumine–enhanced images all showed the central scar with delayed, persistent enhancement to varying degrees. The frequent and characteristic MR findings are summarized in Table 1.

Histopathologic Findings
The gross pathologic features in the 17 totally resected fibrolamellar HCC tumors are listed in Table 2. The mean maximum tumor diameter was 13 cm (range, 3–27 cm). Most had well-defined, lobulated borders and were predominantly of rubbery-to-firm consistency with white-tan color on the surface (Figs 1, 2, 4, 5). The cut surface of the tumors consisted of heterogeneous soft red-brown parenchyma with a more central zone of irregular stellate or bandlike firm white-tan fibrosis (Figs 1, 2, 4, 5). Fibrosis was seen in all cases, and a distinct central scar and radiating fibrous septa were detected in 13 of 17 cases (76%). Hemorrhage was seen in only a few cases and necrosis in less than half, usually as foci with diameter of only a few centimeters (Fig 2). Calcification was recognized in three cases, although no special attempts were made to identify calcific foci.

In all cases in which lymphadenopathy was identified at CT and the patients underwent surgery, attempts were made to completely resect the hepatic tumor along with the involved lymph nodes. Metastatic deposits were present in all resected nodes that were submitted for evaluation.

	DISCUSSION

TOP Abstract Introduction MATERIALS AND METHODS RESULTS DISCUSSION References

 
Our experience with a large number of patients with both modern cross-sectional imaging and histopathologic findings at a single institution reaffirms some prior statements and assumptions about fibrolamellar HCC and calls others into question.

As noted by others (1–6), fibrolamellar HCC has specific and distinctive histopathologic features that clearly distinguish it from conventional HCC in most cases. In addition to the 17 cases in this series with complete resection and pathologic findings in the tumors, we have also reviewed more than 100 cases of fibrolamellar HCC with imaging evaluation or initial surgery at another institution and histologic specimens that were reviewed at our medical center upon referral for further treatment (Ichikawa T, unpublished data, 1999). The microscopic criteria are characteristic, including the diffuse fibrous stroma comprising fibrolamellar bands of collagen and fibrocytes arranged in a lamellar pattern and in delicate bands between nests of tumor cells.

We found that both CT and MR images demonstrated frequent and characteristic features that correlated well with the gross pathologic features of fibrolamellar HCC. Fibrolamellar HCC can be characterized as a large mass in a noncirrhotic liver that occurs primarily in young adults (mean age, 28 years), with no clear sex predominance. The mass is sharply defined (77%) and lobulated (83%) and was often centered in the left lobe (65%), though rarely limited to a single hepatic segment. At CT, the tumor was heterogeneous (90%) with calcification seen in 68%, almost always (95%) within the central scar. A central scar was recognized in 71% of cases, typically stellate (82%) and accompanied by radial septa (95%). Although the tumor itself was heterogeneous, large areas of necrosis or hemorrhage were seen infrequently, and macroscopic fat was never seen. Fibrous tissue could be recognized within the scar and radial septa, which demonstrated persistent enhancement on contrast-enhanced CT and MR images obtained 10–20 minutes after contrast material administration.

Certain imaging features that we encountered are frequently dependent on imaging protocols specifically designed to optimize detection and characterization of liver masses. All of our CT examinations included bolus power injection of contrast media at rates of 2–5 mL/sec: The slower rate of 2–3 mL/sec was used in studies performed with conventional transverse CT scanners, and a rate of 4–5 mL/sec was used as part of a multiphasic helical CT examination. The nonenhanced images were ideal and in some cases essential to recognition of the important feature of central calcification. The hepatic arterial phase images allowed us to recognize the heterogeneous hypervascularity of these tumors (80%). Ten- to 20-minute delayed-phase images were not obtained routinely, but they facilitated recognition of fibrous scars within the tumor, which were seen as areas of delayed persistent enhancement. Although portal venous imaging (60–70-second delay) was performed in all cases in our series, and it is the most common sequence for contrast-enhanced CT in almost all practices, it was actually the least useful for the purpose of identification of distinctive features of fibrolamellar HCC, such as calcification, scar, hypervascularity, and even the size of the tumor (Fig 3).

MR images demonstrated many of the same radiologic-pathologic correlates as did CT scans and could be considered competitive rather than complementary in most cases. Owing to its improved sensitivity to low contrast resolution, MR imaging demonstrates the central scar and fibrous septa even more reliably than does CT, whereas CT is better able to demonstrate calcifications. If CT were performed without delayed imaging, the central and radiating scars could be mistaken for necrotic areas (more characteristic of conventional HCC or metastases), whereas the MR findings correctly suggested solid tissue, which was hypointense on both T1- and T2-weighted images (Figs 2, 3). Other features were demonstrated about equally well with both modalities.

Whether liver-specific MR contrast media will aid detection or characterization of fibrolamellar HCC remains uncertain. Tumor enhancement was seen in one of three fibrolamellar HCC cases with mangafodipir, an MR ligand with selective uptake and enhancement of hepatocytes, but not in the remaining two cases. Similar variability of mangafodipir uptake in conventional HCC has been reported (18), which is apparently related to the degree of histologic differentiation within the tumor.

We believe that a combination of clinical and CT or MR features should allow confident diagnosis of fibrolamellar HCC in most cases and that fibrolamellar HCC can be distinguished from other hypervascular liver masses. Many investigators have stated or speculated that fibrolamellar HCC may closely resemble the benign masses of FNH or hepatic adenoma (7–17). All occur predominantly in younger individuals without underlying cirrhosis. However, FNH and adenomas tend to show homogeneous hypervascular enhancement on hepatic arterial CT or MR images, unlike the heterogeneous enhancement seen in 90% of cases of fibrolamellar HCC (Figs 1, 4, 5). The attenuation or signal intensity of FNH and adenoma is typically similar to that of surrounding liver tissue at CT or MR imaging, respectively, on nonenhanced, portal venous, and delayed-phase images (19–21). Adenomas often demonstrate spontaneous hemorrhage (rare in fibrolamellar HCC) and may have imaging features of focal fat content, which was not encountered previously in cases of fibrolamellar HCC (19). Calcification was seen in almost 70% of cases of fibrolamellar HCC but was rare in FNH or adenoma. Finally, a central scar is commonly seen in FNH, but our experience and that reported by others (7,19,21) suggests that the scar is almost always substantially smaller (as a fraction of the whole tumor) than that typical of fibrolamellar HCC. The small central scar in FNH has been reported (7,19,21) to be hyperintense on T2-weighted images, whereas the larger scar in fibrolamellar HCC is hypointense on T1- and T2-weighted images, in our experience. Of course, rare exceptions may be encountered, such as one case report of a small calcified scar in FNH (22) and another case report of a scar in fibrolamellar HCC that was hyperintense on T2-weighted images (16). Beyond the morphology of the tumor itself, CT or MR images may demonstrate metastatic lymphadenopathy (65% of our cases of fibrolamellar HCC) that would clearly indicate the malignant nature of this tumor.

Large cavernous hemangiomas often have a central scar or necrosis and rarely have central calcification. At CT, however, the nonnecrotic portion of the hemangiomas will be isoattenuating to blood vessels on nonenhanced CT scans and will show progressive nodular or cloudlike enhancement, which is isoattenuating to blood vessels on scans obtained during all phases of enhancement. MR features of hemangiomas, especially marked hyperintensity on T2-weighted images and nodular peripheral contrast enhancement, also allow confident diagnosis and should not be confused with the MR features of fibrolamellar HCC.

In comparison with fibrolamellar HCC, conventional HCC usually has substantially different clinical, laboratory, pathologic, and radiologic features (23–25). Conventional HCC usually affects an older population with underlying chronic liver disease. Central scar, fibrosis, and calcification are rare in conventional HCC, whereas necrosis, hemorrhage, and focal tumor fat are all much more common than is seen in fibrolamellar HCC. Conventional HCC may be unifocal (often encapsulated), multifocal, or diffuse, and it frequently invades portal and hepatic veins. Lacking fibrous tissue, conventional HCC rarely demonstrates delayed persistent enhancement but rather becomes even more hypoattenuating to surrounding liver on delayed images (10 minutes) than on portal venous images (26). Rare sclerosing HCCs could simulate the fibrotic features of fibrolamellar HCC (27). Peripheral cholangiocarcinoma is another primary hepatic malignancy that has abundant fibrous tissue or stroma that may demonstrate delayed contrast enhancement at CT (28). A history of primary sclerosing cholangitis, lack of a stellate central scar, and lack of calcification are features that would favor peripheral cholangiocarcinoma over fibrolamellar HCC. Tumor markers, such as -fetoprotein and des--carboxy-prothrombin, or DCP, are usually elevated in conventional HCC but normal in fibrolamellar HCC (29,30).

A recent article by Winston et al (31) describes MR findings of conventional HCC in cirrhotic and noncirrhotic livers. They describe large, solitary tumors with a central scar in HCC as being more common in noncirrhotic livers, findings that might simulate those seen in fibrolamellar HCC. Although some cases may remain difficult to distinguish, we believe this should be uncommon. The central scars described by Winston et al were frequently hyperintense on MR images and probably represented a variety of causes (necrosis, edema, hemorrhage). Moreover, their patient population with conventional HCC differs substantially from ours with fibrolamellar HCC. Their population was older (mean age, 65 years), serum tumor markers were often elevated, and chronic disease was seen in the noncirrhotic livers.

Distinction of fibrolamellar HCC from metastasis is usually not difficult. Liver metastases usually accompany a known primary tumor and affect an older population. Metastases are often multiple with areas of necrosis or a target appearance and may have other distinctive features, such as amorphous fine calcification in mucinous adenocarcinoma metastases (23). Most metastases are hypovascular and lack a central scar or prominent fibrosis.

We believe that our experience with a relatively large number of patients with fibrolamellar HCC has allowed us to recognize patterns or combinations of imaging features that correlate well with the distinctive pathologic features of this tumor and that allow for confident diagnosis in most cases. To test this belief, we are conducting a blinded interpretation of a large number of cases of proved fibrolamellar HCC, adenoma, FNH, and conventional HCC.

	Footnotes

 
Abbreviations: FNH = focal nodular hyperplasia HCC = hepatocellular carcinoma

Author contributions: Guarantors of integrity of entire study, T.I., M.P.F.; study concepts and design, T.I., M.P.F., L.G.; definition of intellectual content, T.I., M.P.F., L.G.; literature research, T.I., L.G.; clinical studies, T.I., L.G.; data acquisition, all authors; data analysis, T.I., M.P.F., L.G.; manuscript preparation, T.I., M.P.F.; manuscript editing and review, all authors.

	References

TOP Abstract Introduction MATERIALS AND METHODS RESULTS DISCUSSION References

 

1. Craig JR, Peters RL, Edmondoson HA, Omata M. Fibrolamellar carcinoma of the liver: a tumor of adolescents and young adults with distinctive clinico-pathologic features. Cancer 1980; 46:372-379.[Medline]
2. Berman MM, Libbey NP, Foster LH. Hepatocellular carcinoma: polygonal cell type with fibrous stroma—an atypical variant with a favorable prognosis. Cancer 1980; 46:1448-1455.[Medline]
3. Vecchio FM. Fibrolamellar hepatocellular carcinoma of the liver: a distinct entity within the hepatocellular tumors—a review. Appl Pathol 1988; 6:139-148.[Medline]
4. Nagorney DM, Adson MA, Weiland LH, Knight CD, Smalley SR, Zinsmeister AR. Fibrolamellar carcinoma. Am J Surg 1985; 149:113-119.[Medline]
5. Sorelde O, Czemiak A, Bradplece H, Bloom S, Blumgart L. Characteristics of fibrolamellar hepatocellular carcinoma: a study of nine cases and a review of the literature. Am J Surg 1986; 151:518-523.[Medline]
6. Wood WJ, Rawlings M, Evans H, Lim CNH. Hepatocellular carcinoma: importance of histologic classification as a prognostic factor. Am J Surg 1988; 155:663-666.[Medline]
7. Rummeny E, Weissleder R, Sironi S, et al. Central scars in primary liver tumors: MR features, specificity, and pathologic correlation. Radiology 1989; 171:323-326.[Abstract/Free Full Text]
8. Wong LK, Link DP, Frey CF, Ruebner BH, Tesluk H, Pimstone NR. Fibrolamellar carcinoma: radiology, management, and pathology. AJR 1982; 139:172-175.[Free Full Text]
9. Friedman AC, Lichenstein JE, Goodman Z, Fishman EK, Siegelman SS, Dachman AH. Fibrolamellar carcinoma. Radiology 1985; 157:583-587.[Abstract/Free Full Text]
10. Francis IR, Agha FP, Thompson NW, Keren DF. Fibrolamellar carcinoma: clinical, radiologic, and pathologic features. Gastrointest Radiol 1986; 11:67-72.[Medline]
11. Gibson AA, Soreide O, Hemingway AP, Carr DH, Blumgart LH, Allison DJ. The radiology of fibrolamellar carcinoma. Clin Radiol 1986; 37:355-358.[Medline]
12. Brandt D, Johnson CD, Stephens DH, Weiland LH. Imaging of fibrolamellar carcinoma. AJR 1988; 151:295-298.[Abstract/Free Full Text]
13. Titelbaum DS, Burke DR, Meranze SG, Saul SH. Fibrolamellar hepatocellular carcinoma: pitfalls of nonoperative diagnosis. Radiology 1988; 167:25-30.[Abstract/Free Full Text]
14. Titelbaum DS, Hatabu H, Schiebler ML, Kressel HY, Burke DR, Saul SH. Fibrolamellar hepatocellular carcinoma: MR appearance. J Comput Assist Tomogr 1988; 12:588-591.[Medline]
15. Soyer P, Roche A, Levesque M. Legmann P. CT of fibrolamellar hepatocellular carcinoma. J Comput Assist Tomogr 1991; 15:533-538.[Medline]
16. Hamrick-Turner JE, Shipkey FH, Cranston PE. Fibrolamellar hepatocellular carcinoma: MR appearance mimicking focal nodular hyperplasia. J Comput Assist Tomogr 1994; 18:301-304.[Medline]
17. Corrigan K, Semelka RC. Dynamic contrast-enhanced MR imaging of fibrolamellar carcinoma. Abdom Imaging 1995; 20:122-125.[Medline]
18. Murakami T, Baron RL, Peterson MS, et al. Hepatocellular carcinoma: MR imaging with mangafodipir trisodium (Mn-DPDP). Radiology 1996; 200:69-77.[Abstract/Free Full Text]
19. Ros PR. Benign liver tumors. In: Gore RM, Levine MS, Laufer I, eds. Textbook of gastrointestinal radiology. Philadelphia, Pa: Saunders, 1994; 1861-1878.
20. Paulson EK, McClellan JS, Washington K, Spritzer CE, Meyers NC, Baher ME. Hepatic adenoma: MR characteristics and correlation with pathologic findings. AJR 1994; 163:113-116.[Abstract/Free Full Text]
21. Mathieu D, Brunston JN, Arouillard J, Pointreau CC, Vasile N. Hepatic adenomas and focal nodular hyperplasia: dynamic CT study. Radiology 1986; 160:53-58.[Abstract/Free Full Text]
22. Caserio-Alves F, Zins M, Mahfouz AE, et al. Calcification in focal nodular hyperplasia: new problem for differentiation from fibrolamellar hepatocellular carcinoma. Radiology 1996; 198:889-894.[Abstract/Free Full Text]
23. Ros PR. Malignant liver tumors. In: Gore RM, Levine MS, Laufer I, eds. Textbook of gastrointestinal radiology. Philadelphia, Pa: Saunders, 1994; 1897-1946.
24. Matsui O, Itai Y. Diagnosis of primary liver cancer by computed tomography. In: Tobe T, Kameda H, Okudaira M, et al., eds. Primary liver cancer in Japan. New York, NY: Springer, 1992; 129-138.
25. Itai Y, Ohtomo K, Matsui O. Diagnosis of primary liver cancer using magnetic resonance imaging. In: Tobe T, Kameda H, Okudaira M, et al., eds. Primary liver cancer in Japan. New York, NY: Springer, 1992; 139-151.
26. Yoshikawa J, Matsui O, Kadoya M, Gabata T, Arai K, Takashima T. Delayed enhancement of fibrotic areas in hepatic masses: CT-pathologic correlation. J Comput Assist Tomogr 1992; 16:206-211.[Medline]
27. Yamashita Y, Fan ZM, Yamamoto H, et al. Sclerosing hepatocellular carcinoma: radiologic findings. Abdom Imaging 1993; 18:347-351.[Medline]
28. Lacomis JM, Baron RL, Oliver JH, Nalesnik MA, Federle MP. Cholangiocarcinoma: delayed CT contrast enhancement patterns. Radiology 1997; 203:98-104.[Abstract/Free Full Text]
29. Chen DS, Sung JL. Serum alpha-fetoprotein in hepatocellular carcinoma. Cancer 1977; 40:779-783.[Medline]
30. Fujiyama S, Morishita T, Hashiguchi O, Sato T. Plasma abnormal prothrombin (des-gamma-carboxy prothrombin) as a marker of hepatocellular carcinoma. Cancer 1988; 61:1621-1628.[Medline]
31. Winston CB, Schwartz LH, Fong Y, Blumgart LH, Panicek DM. Hepatocellular carcinoma: MR imaging findings in cirrhotic livers and noncirrhotic livers. Radiology 1999; 210:75-79.[Abstract/Free Full Text]

This article has been cited by other articles: